Research On Grey Nonlinear Energy Forecasting Model Driven By System Structure

Energy is an important lifeline that affects the economic development of all countries in the world.Objective and accurate energy consumption forecast can provide important reference and leading indicators for the government to implement economic policies and energy development strategies.In recent years,China’s energy consumption shows a steady growth trend.Shandong Province,Guangdong Province and Jiangsu Province are the three major energy consumption provinces in China.The energy supply and demand of each province are different.Among them,Shandong Province has the largest energy consumption,but its economic development is relatively slow,so it is necessary to improve the energy utilization rate.As a province with rapid economic development in China,Guangdong’s energy demand is also growing rapidly.Jiangsu province consumes a large amount of energy,but it is short of energy resources,so it depends on other provinces.Therefore,it is of great significance to accurately predict the energy consumption of large provinces,which is conducive to the formulation of provincial development plans.The research on the rapid growth trend of consumption will lay a foundation for the protection of people’s normal life and rapid economic development.In order to accurately predict the energy consumption of each province,the related work is as follows:Based on the classical grey GM(1,1)model,a new extended model GM(1,1,v(t))is proposed.When v(t)is a different type of function,the model can consider linear and nonlinear factors.GM(1,1,v(t))can be expressed as different types of grey model.In addition,through theoretical verification,the change of number multiplication does not affect the prediction accuracy of the model,which lays a theoretical foundation for complex data preprocessing.Secondly,using the total energy consumption of Shandong Province,Guangdong Province and Jiangsu Province from 2006 to 2018 as the research object,eight models are used to fit the test,and the most suitable model for the energy consumption data of the province is found.The results show that DGM(1,1)is the best in Shandong Province,GM(1,1)is the best in Guangdong Province,and Grey Verhulst Model is the best in Jiangsu Province.On account of the complexity and uncertainty of the energy system,this paper establishes a novel flexible grey multivariable model by introducing a power exponential term,a linear correct term and a random disturbance term.The novel model has the advantages in capturing the dynamic characteristics of the energy system,also it can be compatible with the classical grey models when the parameters are in special situations.Additionally,to further promote the prediction performance of the novel model,the grey wolf optimizer is employed to determine the power parameters of the model.To demonstrate its efficacy,the proposed model is utilized to predict the energy consumption of three major province in China,and the simulation and prediction results of the novel model are compared with those provided by diversified competing models.The results illustrated that the novel model is superior to other competing models,offering more accurate and reasonable performance.Based on the results,several proposal for energy development are put forward for decision-makers.